Principles of Large-C Research

Question 1

What is a case? Give an example

Answer 1

A case is a spatially and temporally delimited phenomenon of theoretical interest. For example, a country, the social media posts by a political party, or a survey respondent. Typically contains multiple observations.

Question 2

How do cases often differ from observations?

Answer 2

A case is often the same as an observation, but observation has to be defined as the LOWEST-LEVEL unit in an analysis, not merely an individual case.

Question 3

What is an observation? Give an example

Answer 3

An observation is the lowest-level unit in an analysis at which a measured variable can only take one value (unit of analysis). For example, a country-year, an individual post on social media, or a survey respondent. May or may not be of theoretical interest.

Question 4

What is a sample? Give an example

Answer 4

A sample is the set of cases or observations that are analysed in a given piece of research. For example, all European countries during a given time period, all social media posts by the Democrats and Republicans, all individuals in the UK.

Question 5

What is a population? Give an example

Answer 5

A population is the set of cases which in combination make up the universe of cases, i.e., all cases. For example, all countries in the world during a given time period, all social media posts by political parties, all individuals in the UK

Question 6

What is population inference?

Answer 6

Population inference is when a researcher infers something they do not know (patterns in a POPULATION) from something that is known (patterns in a SAMPLE)

Question 7

What instances may sampling biases occur in?

Answer 7

Sampling bias may occur if a particular sample may be skewed a certain way due to how it is selected or if a sample may give different conclusions based on what is included.

Question 8

What issues are there with case selection based on the dependent variable? When can this be acceptable?

Answer 8

Selection of cases based on the dependent variable means you are selecting cases based on OUTCOME, precluding any attempt to look for a causal relationship (BIASED CAUSAL INFERENCES). This can be acceptable in exploratory research and to generate new theory.

Question 9

What can lead to sampling biases?

Answer 9

Sampling biases may be caused by distorted sampling frames, selection on the dependent variable, survivorship bias, non-response bias, self-selection into the sample, etc.

Question 10

What practices should be avoided to prevent sampling biases?

Answer 10

• Cherry-picking cases that are KNOWN to support a hypothesis - sample must show variation in characteristics
• Performing an inductive study and using same set of cases to test the resulting theory and hypotheses
• In large-C research: selecting cases based on the dependent variable

Question 11

What strengths are there of Large-C research?

Answer 11

Large C research increases the potential for generalisability (given that random sampling means that population inferences can be made) and increases the ability to identify causal effects.

Question 12

What weaknesses are there of Large-C research?

Answer 12

Large-C studies often overlook the individual circumstances of cases, they are less useful for indicative research and are not as valuable for interpretivist research due to their ‘thin’ form of analysis.

Question 13

What are the key principles of case selection in Large-C studies?

Answer 13

Selections of cases must involve a sample that is…
- Large enough sample to enable robust statistical inferences
- Sample that is representative of population

Question 14

Explain total population sampling. What are the benefits of this?

Answer 14

Total population sampling looks at all cases in a population. This produces high external validity, and discards arguments on whether a sample is representative.

Question 15

Explain simple random sampling

Answer 15

Simple random sampling is a form of probability sampling where all cases are drawn from the population with the same probability

Question 16

What is the law of large numbers? How is this navigated through with regards to simple random sampling?

Answer 16

The law of large numbers holds that the larger a sample, the more likely it is that this will mirror general trends in the general population. Thus, making simple random samples larger, it is more likely it will represent the population.

Question 17

Explain random statified sampling

Answer 17

Stratified random sampling involves grouping populations into relevant strata (subclasses), and then drawing cases at random from these strata to ensure that important groups are adequately represented.

Question 18

What is non-probability sampling? Why is this often used?

Answer 18

Non-probability sampling involves any sampling method where the probability of a selection is unknown. It is often used due to being cheaper and more convenient than random alternatives.

Question 19

What issues are there with non-probability sampling?

Answer 19

Non-probability sampling means that insights are difficult to generalise into the full population, and can easily give rise to sample selection biases.